Neuronal loss occurs in the hippocampus and other related brain areas during the development of Alzheimer's disease. Since the anatomic integrity of the hippocampus has been shown in both animals and humans to be a critical for memory, neurodegeneration of this brain structure may form the basis for memory losses in patients with dementia of the Alzheimer type (DAT). While cholinesterase inhibitors have been shown to ameliorate memory losses in patients with DAT, not all patients improve with treatment. Unfortunately, because we cannot predict how patients will respond, many patients with DAT are needlessly exposed to the substantial costs and side effects of ineffective therapy. Hippocampal neurons receive cholinergic afferents from the nucleus basalis of Meynert. In this project, we propose to test the hypothesis that the integrity of such neurons determines the outcome of cholinesterase inhibitor therapy, using two experimental approaches: 1) predicting the outcome of donepezil treatment in DAT subjects using in vivo neuromorphometry of the hippocampus and related brain structures, and 2) examining the effects of cholinomimetics on memory-related tasks in animals subjected to pharmacologic, excitotoxic and metabolic lesions of the hippocampus. While the first experiment has direct clinical application, the animal model experiments are critical for determining the neurobiological basis of any relationship between hippocampal neurodegeneration and treatment responsiveness in patients with DAT. Specific Aim 1: Fifty subjects with very mild-to mild DAT will be recruited from the Memory Diagnostic Center at Washington University to receive a high resolution MR scan and open label treatment with donepezil for two years. Clinical assessments will occur every three months. High resolution MR scans will be analyzed with high dimensional brain mapping tools to assess the volume and shape characteristics of the hippocampus and related brain structures. Specific Aim 2: The efficacy of cholinomimetic drugs will be studied in three animal models of hippocampal neurodegeneration: 1) pharmacological blockade of NMDA receptors using MK-801; 2) excitotoxic lesions of the hippocampus: and 3) transgenic mice expressing mutant forms of human genes encoding amyloid precursor protein and presenilin 1. In each of these animal models, memory-related deficits occur that are similar to memory deficits in DAT patients.